The primary objectives of the proposal are: 1) To develop an objective Behavior Observation Scale (BOS) which will differentiate autistic from normal and mentally retarded children (30-60 months of age); 2) To differentiate subgroups of autistic children along various symptom axes; 3) To establish a quantitative continuum of severity within each symptom subgroup; 4) To provide an objective means of describing children who serve as subjects in neurophysiological and neurobiochemical research; 5) To periodically re-evaluate autistic children and thus trace the natural course of the disease. In order to accomplish these objectives, normal, autistic and retarded children will be observed in a structured setting and their behaviors counted on the Behavior Observation Scale (BOS). Data obtained on the BOS will be subjected to a discriminant analysis which will: 1) Differentiate among groups of children; 2) Divide children according to symptom subclusters and 3) Provide a quantitative continuum of severity in each subcluster. These data will be cross-correlated with data obtained from developmental history, psychiatric interviews, clinical, psychometric and language evaluations.